Terminology Consistency in English-Arabic Women's Garment Translations on SHEIN and Amazon in KSA: A Comparative Analysis

The e-commerce industry in Saudi Arabia is growing rapidly, making this research essential. Inconsistent translations may lead to consumer misunderstanding, lost sales, and a reduction in platform trust, highlighting a significant flaw in localization methods. This study evaluates the consistency of English-to-Arabic translations of women's garment terminology on the e-commerce platforms SHEIN and Amazon KSA. The research employs qualitative, product-focused comparative content analysis. Sixty product descriptions (tops, dresses, and bottoms) were chosen at random from SHEIN and Amazon KSA. Every description was analyzed and categorized according to 12 predefined categories, including style coherence, borrowing methods, and language diversity. Original product descriptions were saved as screenshots, and manual coding was verified for consistency to ensure the accuracy of the data. The findings revealed that SHEIN exhibits a higher frequency of translation inconsistencies compared to Amazon, directly addressing the first research question by identifying discrepancies such as inconsistent terminology (e.g., "pants" translated as ([بنطلون/بنطال/سروال] {dir=”rtl”}) and conceptual mismatches. These results addressed the second research question by demonstrating how limited databases of Arabic fashion terminology and structural differences between English and Arabic pose challenges for machine translation models, especially when considering dialectal and grammatical nuances. This study contributes to the fields of e-commerce translation and localization research by (1) empirically analyzing the limitations of machine translation in Arabic fashion terminology, (2) proposing practical improvements such as standardized glossaries and streamlined post-editing processes, and (3) highlighting the need for Scopus-based practices to ensure that translations meet user needs.